The transparent cornea is made up of an anterior stratified epithelium, a collagenous stromal matrix containing fibroblasts called keratocytes, Descemet's membrane and a single-layered endothelium. We have put forth the refracton hypothesis, which speculates the cornea is more similar to the lens than previously believed. The cornerstone of this hypothesis is that the exceptionally abundant intracellular proteins in the cornea function like the multifunctional crystallins of the lens. Our current corneal research addresses the molecular basis for corneal-specific gene expression and the nature of the roles of the abundant, intracellular corneal proteins. For corneal-specific gene expression, we are continuing to investigate the mouse aldehyde dehydrogenase 3 (ALDH3), rabbit ALDH1 and zebrafish gelsolin promoter in both transfection and transgenic mouse experiments. Detailed studies on the mouse ALDH3 promoter activity are just beginning. We showed that the corneal epithelial cells of Pax6+/- Small eye mice carrying two different mutations (SeyNeu or SeyDey) are poorly stratified and have reduced amounts of ALDH3, TKT and K12. This indicates that high expression of these proteins in the corneal epithelium require proper amounts of Pax6 during development. The 4 kb ALDH3 promoter/CAT reporter gene, shown earlier to be expressed specifically in corneal epithelial cells of transgenic mice, was stimulated 10-fold in co-transfection experiments with Pax6 in Cos 7 cells. Thus, Pax6 may have a role in corneal ALDH3 expression in mice as it does in lens-specific promoter activity. We showed earlier in collaborative studies with Dr. James Jester (Southwestern Medical Center, Dallas, Texas) that ALDH1 is the principal protein in rabbit corneal keratocyes. It is also a major protein of the rabbit corneal epithelium. The rabbit ALDH1 gene has been characterized and truncated fragments of its promoter are being tested in transfection and transgenic mouse experiments. A 1 kb promoter fragment is active in cornea and, especially, in retina of transgenic mice; an approximate 3 kb promoter fragment is presently being tested for comparison. Since there is a paucity of corneal cell lines to use in transfection experiments, we have continued to examine the mouse corneal epithelial A6(1) cell line from the 14 day-old immortomouse that we reported last year. Since the immortomouse carries a g-interferon-inducible, temperature-sensitive SV40 T-antigen, the A6 cells grow faster at 33o C in the presence of ?-interferon than at 33 degrees in the absence of ?-interferon. The switch to 37 degrees is accompanied with an upregulation of ALDH3 and, to a lesser extent, cytokeratin K12. The expression of a-enolase, a preferentially expressed marker for basal epithelial cells, decreases in the A6 cells at 37o C in the absence of ?-interferon. Although the A6 cells undergo some changes resembling corneal differentiation in culture, the expression levels for ALDH3 and K12 are low at best, and the cells do not stratify when cultured on amniotic membrances. Nonetheless, we anticipate they will be useful for transfection studies. We reported last year that gelsolin, an actin binding protein, comprises approximately half of the water-soluble protein of the zebrafish corneal epithelium, making it a putative corneal crystallin in this species. Current in situ hybridization experiments showed that gelsolin undergoes a burst of expression at the blastula stage of zebrafish development, becomes expressed in the notocord, and then becomes concentrated in the eye. Microinjection of gelsolin-specific morpholino oligonucleotides in 2-cell stage zebrafish embryos selectively affected eye and tail development, indicating a developmental role. In addition, gelsolin promoter/reporter gene constructs are being made for study in transfection and transgenic mouse experiments. Last year we reported the creation of knockout mice in order to explore the functional role of ALDH3 and TKT in the cornea. ALDH3 null mice were viable. Surprisingly, no abnormal phenotype was observed in these mice even after more than a year. The cornea is clear when examined with the slit lamp and appears normal by light and electron microscopy. There is no evidence for the appearance of a compensatory protein. Protein concentration per mg/DNA of the corneal epithelium appears to be approximately half that of normal mice. TKT null mice died in the early cleavage stages of the embryo. However, heterozygote TKT mice were viable and showed no corneal abnormalities. Unexpectedly, the TKT mice were on average about 30% smaller than wild type mice. The liver, testis, ovary and, especially, fat tissue were proportionately smaller in the TKT+/- mice, however the size of the brain and kidney appeared normal. It appears that adipose deficiences are largely responsible for the reductions in size of the TKT heterozygous mice. TKT heterozygote females gave small litters when mated with wild type males, and we have some evidence for reduced sperm motility in the TKT heterozytote males. These unexpected dose-dependent TKT phenotypes may have clinical relevance. We are attempting to obtain transgenic mice that express the Cre gene specifically in corneal epithelial cells in order to perform corneal-specific gene knockouts. We tested a mouse ?339/+44 aB-crystallin promoter/Pax6 enhancer/lacZ transgene since we had shown that this aB-crystallin promoter fragment is expressed specifically in lens and corneal epithelial cells, and others have shown that the Pax6 enhancer directs activity of a heterologous basal promoter to the same two tissues in transgenic mice. Six transgenic mouse lines have been obtained. One expresses lacZ specifically in the adult retina and one expresses in the adult retina and cornea; developmental expression has not been investigated yet in either line. Four other lines remain to be examined. We have made cDNA libraries of whole cornea and of the isolated endothelium/stroma of 6-week-old mice. Approximately 1000 clones have been sequenced from each library. As expected the most abundant clones from the total library are for K12 keratin and ALDH3. Recently we have made a SAGE library from 6-week-old mouse corneas and have identified approximately 2500 separate expressed genes. So far the SAGE and total cDNA libraries are in good agreement. We have also used subtraction hybridization to identify regulatory genes used in corneal development. A cDNA (called CORE for corneal epithelium) that was enriched in the newborn mouse corneal epithelium was isolated and characterized. CORE appears to be expressed in multiple tissues and has two LIM domains and a C-terminal zinc finger.